You are here: Administrators > Content Library > Data Instances > Create Data Instance via upload

Create Data Instance via upload

If you chose "Create via Upload" as described in the Data Instances article, you will go through the steps defined in this article.

With this option, you are able to upload an excel file and create multiple Data Instances.

Once you've clicked "create via upload", you are taken to the next step where you are prompted to upload a file. as shown below.

  1. Drag and drop your file or click "browse" to select the file you wish to upload
  2. Once you have selected the file, the system processes it.

If the file is a valid format, the system will take you to the field Mapping step as described below.

Field Mapping

Here ContentWelder will ask you to map the columns from your file the selected Data Model fields.

  Name Description
1 File Columns

On the left hand side, you will see the list of columns from your file.

The first Row in your file is the column Header

So in this case the first row in the uploaded spreadsheet contained fields named "Courtesy, First Name, Last Name, Company etc."

Beneath each column header, the second row of data is displayed.

2 Data Model Available Fields

On the right hand side, you will see the list of available Data Model fields

Fields highlighted in red are required

3 Previous Click to go back to the previous step and choose a different way of uploading content.
4 Finish

Click to complete your mapping

If you click before mapping all required fields, the following error is displayed.

5    

Data Type Data Validation

Since the creation of data instance based on files happens behind the scenes, errors are as evident as when you enter content manually. below is a description of the data type validation logic and how it works with file uploads.

 

Validating Required fields

Required fields are the first to be checked for both their existing as well as for data type validation.

Note: Remember, in a spreadsheet just because the first row has all the required fields filled out, doesn't mean row 423 does as well

  1. If a required field is missing
    1. The row is ignored and the system moves onto the next
  2. Or if a field's data type doesn't match the one defined in the Data Model
    1. The row is ignored and the system moves onto the next
  3. Or if a Data Model has a "length" defined and the field violates this length
    1. The row is ignored and the system moves onto the next.

In short, required field must be valid otherwise the entire row is ignored.

 

Validating non Required fields

  1. If a field's data type doesn't match the one defined in the Data Model
    1. The column is ignored and will not be updated
  2. Or if a Data Model has a "length" defined and the field violates this length
    1. The column is ignored and will not be updated